ABSTRACT
Embryos from XO female mice begin development with half the activity levels of an enzyme (HPRT) coded for by a gene on the X chromosome, compared with embryos from XX females. Groups of unfertilized eggs and individual embryos at the 8-cell, morula and blastocyst stages were assayed for HPRT activity. An autosomally coded enzyme (APRT) was assayed simultaneously in the same reaction mix as a control. There is a substantial increase in HPRT activity by the 8-cell stage. However, the mean activity of HPRT in embryos of XO mothers remains half that in embryos of XX mothers. This suggests a signifi- cant maternally inherited component of HPRT activity in 8-cell embryos. By the 9- to 16-cell morula stage the HPRT activities in the two groups of embryos become similar due, pre- sumably, to a transition to embryo-coded activity; HPRT activities in individual morulae from XX mothers show a bimodal distribution consistent with the hypothesis that both X- chromosomes are active in XX embryos at this stage.
INTRODUCTION
Previous work (Monk & Kathuria, 1977) has shown that XX and XY embryos appear to be equivalent with respect to X-linked HPRT (hypoxanthine phosphoribosyl transferase, E.C. 2.4.2.8) activity at the early 8-cell and the blastocyst stages. The distribution of HPRT activities for large numbers of individual embryos did not correspond to two populations (XX and XY) differing by a factor of two, which might be expected if XX embryos had twice the HPRT activity of XY embryos. We concluded that some form of dosage compensation for the Y-linked HPRT activity was operating at these developmental stages. Although there is evidence that HPRT activity is embryo-coded in morulae and blastocysts on the fourth day of pregnancy (Epstein, 1972), it remained possible that the activity at the 8-cell stage was maternally derived.
In this work we have extended Epstein’s earlier studies (Epstein, 1972) to an investigation of levels of HPRT activity in single embryos throughout preimplantation development. This enzyme is coded by the X chromosome, and because both X chromosomes are active during oogenesis (Epstein, 1969, 1972; Kozak, McLean & Eicher, 1974) embryos of XO mothers begin development with half as much HPRT as do embryos of XX mothers. The time at which activities in. the two groups of embryos become similar presumably reflects the onset of expression of the embryonic HPRT gene. We show here that there is a considerable increase in HPRT activity up to the 8-cell stage although the activity levels in AO-derived embryos remain half those in AA-derived embryos. As a control we show that activities of an autosomally-coded enzyme, APRT (adenine phospho-ribosyl transferase, E.C. 2.4.2.7), remain similar in the two groups of embryos at each stage, while also rising with advancing development. Beyond the 8-cell stage HPRT levels become the same in XX- and AO-derived embryos as the embryo-coded HPRT activity predominates. The distribution of HPRT activities in individual 9- to 16-cell embryos from XX mothers is bimodal with a twofold separation of the modes. This result suggests that both X chromosomes are active in female XX embryos prior to implantation.
MATERIALS AND METHODS
Mouse embryos were obtained from randomly bred MFI females (Olac 1976 Ltd) or from XX and XO females (genotypes Ta/ + and + /O; kindly provided by Mary Lyon, MRC Radiobiology Unit, Harwell). The mice were super-ovulated by intraperitoneal injection of 5 i.u. of PMS (pregnant mare serum) and, 48 h later, 5 i.u. of HCG (human chorionic gonadotrophin) and then immediately caged with MFI males. Eggs were isolated from the oviducts the morning following mating, 2-cell, and 8-cell and 9- to 16-cell embryos from the oviducts on the second and third day of pregnancy respectively, and early and late blastocysts from the uterus on the fourth and fifth day of pregnancy respectively. The medium used for collection of eggs and embryos (PB1) and the methods of isolation of embryos are described in Monk & Ansell (1976).
The enzymes HPRT and APRT in embryo extracts were simultaneously assayed in the same reaction mix (McBurney & Adamson, 1976; Monk & Kathuria, 1977). XX- or ATI-derived eggs, 8-cell embryos and 9- to 16-cell morulae and blastocysts were transferred to 10 μ1 Drummond caps, either singly or in groups of five to ten, in less than 5 μ1 of medium. The microcaps were sealed and extracts prepared by freeze-thawing in liquid nitrogen three times, followed by centrifugation for 10 min. The supernatant from each embryo extract was added to a 50 μ1 reaction mixture containing sodium phosphate buffer (35 mM, pH 7-4), magnesium chloride (5 mM), phospho-ribosyl pyrophosphate (2 mM), [3H]guanine (10 μM, specific activity 84 mCi/mM) and [14C]-adenine (10 μM, specific activity 58 or 61 mCi/mM). The reactions were carried out at 37 °C for times ranging between 2 and 4 h and stopped by the addition of 1 ml volumes of cold lanthanum chloride (0·1 M) containing adenine and guanine (1 mM). The linearity of the reactions with times of incubation for the embryos examined had previously been established (Monk & Kathuria, 1977). Corrections for spill-over of [3H] counts into the [14C] channel, and vice versa, were determined in reactions containing embryo extract and [3H] or [14C] alone respectively. The level of counts ranged from three times ([14C] counts for 8-cell embryos) to ten times ([3H] counts for blastocysts) the counts in replicate reproducible blank samples containing 5 of supporting medium but no embryo extract. Counting efficiencies for [3H] and [14C] were determined by spotting 2 μI of reaction mixture directly onto filters. Some day-to-day variability was unavoidable, e.g. the calculated enzyme activities shown in Figs. 1, 4 and 5 are relatively low, but the data in each experiment are internally consistent and this variability does not affect the interpretation of the results.
Activities of HPRT and APRT in developing MFI embryos. Fertilized eggs, 2-cell embryos, 8-cell embryos and morulae, and early and late blastocysts were isolated on the first, second, third, fourth and fifth days of pregnancy respectively. Extracts from batches of five to ten embryos were assayed for HPRT and APRT. Assays were performed for 3 h as described in Materials and Methods. Results are plotted against the number of hours after mating, assuming mating occurs at midnight.
Activities of HPRT and APRT in developing MFI embryos. Fertilized eggs, 2-cell embryos, 8-cell embryos and morulae, and early and late blastocysts were isolated on the first, second, third, fourth and fifth days of pregnancy respectively. Extracts from batches of five to ten embryos were assayed for HPRT and APRT. Assays were performed for 3 h as described in Materials and Methods. Results are plotted against the number of hours after mating, assuming mating occurs at midnight.
For single embryo assays increased accuracy is achieved by the expression of the results as ratios of HPRT:APRT activities where both enzymes are measured simultaneously in the same reaction mix. This eliminates variance due to variable recovery of embryo extracts.
RESULTS
Throughout these experiments we have assayed both HPRT, coded by the X chromosome, and as a control, APRT which is coded autosomally. Fig. 1 shows the increase in activities of HPRT and APRT during preimplantation development. The two enzymes were measured simultaneously in a given batch of embryos at a particular stage of development. Essentially similar results were previously published by Epstein (1970) who used high voltage paper electro-phoresis to measure each enzyme separately in extracts of embryos at comparable stages.
Because both X chromosomes are active during oogenesis, embryos from XX mothers would have twice the levels of HPRT compared with embryos from XO mothers for as long as the activity observed was maternal in origin. Fig. 2 shows the twofold difference in HPRT activities in unfertilized eggs from XO and XX females, thus confirming the results of Epstein (1972). The ratios of activities of HPRT: APRT in the XO and XX eggs are different by a factor of three due to rather higher APRT activities in the XO eggs in these particular experiments.
Activities of HPRT, APRT, and ratios of activities HPRT: APRT in batches of eggs (five to ten eggs per batch) from XX and XO (hatched squares) females. Each square represents a separate batch of eggs. The mice were superovulated and unfertilized eggs collected at approximately 21 h following HCG injection. Assays were performed for 4 h. Mean activities (p-mole per hour per egg± S.E.), XXeggs: HPRT 0·38 ± 0·01, APRT 0·16 ± 0·005, HPRT/APRT 2·4 ± 0·07; XO eggs: HPRT 0·17 ± 0·01, APRT 0·21 ± 0·01, HPRT/APRT 0·8 ± 0·06.
Activities of HPRT, APRT, and ratios of activities HPRT: APRT in batches of eggs (five to ten eggs per batch) from XX and XO (hatched squares) females. Each square represents a separate batch of eggs. The mice were superovulated and unfertilized eggs collected at approximately 21 h following HCG injection. Assays were performed for 4 h. Mean activities (p-mole per hour per egg± S.E.), XXeggs: HPRT 0·38 ± 0·01, APRT 0·16 ± 0·005, HPRT/APRT 2·4 ± 0·07; XO eggs: HPRT 0·17 ± 0·01, APRT 0·21 ± 0·01, HPRT/APRT 0·8 ± 0·06.
The activities of HPRT, APRT, and their ratios in single 8-cell embryos from XX and XO (hatched squares) mothers are shown in Fig. 3. Although the autosomally coded APRT activities are similar in both groups of embryos, the H PRT activities, and hence the HPRT : APRT ratios, are approximately two-fold lower for XO embryos, thus reflecting the maternal origin (XX or XO) of the embryo (Fig. 2). In addition the HPRT activities in the XT-derived 8-cell embryos in Fig. 3 are consistent with a bimodal distribution (modes approximately 1·5:1) which might indicate the separation of XX and XY embryos due to an effect of X-chromosome dosage on embryo-coded HPRT activity. Similar bimodal (approximately 1·5:l) distributions were obtained in four other experiments with XX-derived 8-cell embryos.
Activities of HPRT, APRT, and ratio of activities HPRT: APRT in single 8-cell embryos isolated from XX and XO (hatched squares) females on the third day of pregnancy at 64 and 68 h, respectively, after HCG injection. Each square represents a value obtained for a separate embryo. The results from two experiments were pooled. Assays were performed for 4 h. Mean activities (p-mole per hour per embryo±S.E.), XX embryos: HPRT 7·5 ± 0·4, APRT 0·22 ± 0·01, HPRT/APRT 35·4 ±l·4; XO embryos: HPRT 4·0 ± 0·3, APRT 0·22 ± 0·01, HPRT/APRT 18·9 ± 1·2.
Activities of HPRT, APRT, and ratio of activities HPRT: APRT in single 8-cell embryos isolated from XX and XO (hatched squares) females on the third day of pregnancy at 64 and 68 h, respectively, after HCG injection. Each square represents a value obtained for a separate embryo. The results from two experiments were pooled. Assays were performed for 4 h. Mean activities (p-mole per hour per embryo±S.E.), XX embryos: HPRT 7·5 ± 0·4, APRT 0·22 ± 0·01, HPRT/APRT 35·4 ±l·4; XO embryos: HPRT 4·0 ± 0·3, APRT 0·22 ± 0·01, HPRT/APRT 18·9 ± 1·2.
To determine when embryo-coded activity began to obliterate maternal differences we examined slightly older embryos, namely morulae at the 9- to 16-cell stage. These were compacted and an attempt was made to include only those where the outline of the morula showed a fourth cleavage division had occurred. Figure 4 shows that XO-derived morulae (hatched squares) now have the same HPRT, APRT and HPRT: APRT ratios as XX-derived morulae. We conclude that HPRT activity is embryo-coded in 9- to 16-cell morulae isolated late on the third day of pregnancy. An analysis of a large number of individual 9- to 16-cell morulae derived from XX-mothers shows a clear bimodal distribution, with an approximate twofold separation of the modes, for HPRT and HPRT: APRT ratios (Fig. 5). The distribution of values for the autosomal enzyme, APRT, is unimodal. The results suggest two populations of embryos, XX and XY, differing by a factor of two with respect to X-chromosome activity.
Activities of HPRT, APRT, and ratios of activities HPRT: APRT in single 9-16 cell morulae isolated from XX and XO (hatched squares) females at 70 and 76 h, respectively, after HCG injection. Assays were performed for 4 h. Mean activities (p-mole per hour per embryo ± S.E.), XXembryos: HPRT 6·2 ± 0·6, APRT 0·16 ± 0·01, HPRT/APRT 37·8 ± 2·9; XO embryos: HPRT 7·1 ± 1·3, APRT 0·25 ± 0·06, HPRT/APRT 32·3 ± 3·3.
Activities of HPRT, APRT, and ratios of activities HPRT: APRT in single 9-16 cell morulae isolated from XX and XO (hatched squares) females at 70 and 76 h, respectively, after HCG injection. Assays were performed for 4 h. Mean activities (p-mole per hour per embryo ± S.E.), XXembryos: HPRT 6·2 ± 0·6, APRT 0·16 ± 0·01, HPRT/APRT 37·8 ± 2·9; XO embryos: HPRT 7·1 ± 1·3, APRT 0·25 ± 0·06, HPRT/APRT 32·3 ± 3·3.
Activities of HPRT, APRT, and ratios of activities HPRT/APRT in single 9-16 cell morulae isolated from MFI females at 74 h after HCG injection. Assays were performed for 4 h. Mean activities (p-mole per hour per embryo ± S.E.) HPRT 1·51 ± 0·07, APRT 0·11 ± 0·003, HPRT/APRT 13·06 ± 0·48.
Blastocysts isolated from XO and XX females on the next (4th) day of pregnancy (Fig. 6) also show equivalent levels of HPRT and APRT activities, but now the separation of presumptive XX and XY embryos is less than twofold.
Activities of HPRT, APRT, and ratios of activities HPRT: APRT in single blastocyst isolated from XX and XO (hatched squares) females on the fourth day of pregnancy at 97 and 99 h, respectively, after HCG injection. Assays were performed for 3 h Mean activities (p-mole per hour per embryo ± S.E.), XX embryos : HPRT 29·0 ± 1·7, APRT 0·48 ± 0·01, HPRT/APRT 59·6 ± 2·7; XO embryos: HPRT 28·4 ± 2·8, APRT 0·57 ± 0·06, HPRT/APRT 58·7 ± 9·7.
Activities of HPRT, APRT, and ratios of activities HPRT: APRT in single blastocyst isolated from XX and XO (hatched squares) females on the fourth day of pregnancy at 97 and 99 h, respectively, after HCG injection. Assays were performed for 3 h Mean activities (p-mole per hour per embryo ± S.E.), XX embryos : HPRT 29·0 ± 1·7, APRT 0·48 ± 0·01, HPRT/APRT 59·6 ± 2·7; XO embryos: HPRT 28·4 ± 2·8, APRT 0·57 ± 0·06, HPRT/APRT 58·7 ± 9·7.
DISCUSSION
Epstein (1972) has previously studied the activities of HPRT in eggs and 2-cell stage embryos, and in morulae and blastocysts on the 4th day of pregnancy, in batches of embryos issuing from XX and XO mothers. Eggs and 2-cell stage XX-derived embryos had twice as much HPRT activity as those from XO mothers, thus reflecting the Xchromosome dosage during oogenesis. There was little, if any, increase in activity from the egg to the 2-cell stage in embryos from either XX or XO mothers. Epstein did not report results for embryos taken on the third day of pregnancy, though on the fourth day of pregnancy the approximate equivalence of HPRT activities in batches of embryos from XO and XX mothers led him to postulate that, at this stage, HPRT activity was embryocoded.
Embryos from XX and XO mothers will have different sex chromosome constitutions (XX and XY, and XX, XY, XO and YO, respectively). The YO embryos are thought to be arrested at the 8-cell stage (Burgoyne & Biggers, 1976). Theoretically, if both X chromosomes had been active in generating the activity observed in groups of blastocysts (Epstein, 1972), then the XO-derived embryos should show a predictable fraction (0-89 or higher) of the HPRT activity of the XX-derived embryos. This fraction is derived from the expectation of four active X chromosomes in three embryos (XX, XO and XY) compared with three active X chromosomes in two embryos (XX and XY). The value would be greater than 0-89 since the proportion of XO embryos from the XO mothers is less than one third (Kaufman, 1972). If X inactivation had occurred, HPRT activity in groups of blastocysts from XX and XO mothers should be equivalent. Epstein could not deduce from his data whether both Xchromosomes were active in preimplantation development.
The application of the above argument to the data in Fig. 3 indicates that HPRT activity in the 8-cell embryos analysed is unlikely to be totally embryocoded. If this were so we would expect the ratio of the mean HPRT activities, or higher if XO plus OY embryos are present to less than 50 per cent of the litter; or
or higher if only one X chromosome is active in XX embryos up to the 8-cell stage. Statistical analysis shows that the ratio of the mean HPRT activities, XO to XX, in Fig. 3 (0·53) is significantly different from 0·67 (P < 0·01).
On the other hand the indications of an approximately T5-fold separation of two populations with respect to HPRT activity in the XX-derived 8-cell embryos in Fig. 3 would suggest that some embryo-coded activity is already present. The data presented in Fig. 3 can be most easily interpreted as resulting from a mixture of approximately equal amounts of maternally inherited and embryo-coded HPRT activity. We previously reported a unimodal distribution of HPRT activities for 8-cell embryos (Monk & Kathuria, 1977). When reexamined at the level of single litters of embryos this earlier data is also consistent with 1·5 fold separation of two populations of embryos. Similar results have been obtained by Kratzer & Gartler (pers. comm.). Since there is a 10- to 20-fold increase in HPRT activity by the 8-cell stage it seems highly likely that some increase has occurred in maternally derived enzyme activity. These results may therefore represent the first evidence in mammalian embryos of active stable maternal messenger RNA functional up to the fourth cleavage in preimplantation development, although the gradual activation of pre-existing maternal precursor protein for HPRT activity cannot be excluded.
An alternative explanation for the lack of a twofold separation of embryos at the early 8-cell stage might be that the onset of expression of the maternally derived HPRT gene occurs earlier than that on the paternally derived X chromosome. This appears unlikely in view of the fact that there is no evidence in Fig. 3 of XX and XY embryos from XO mothers with equivalent activity to those from XX mothers. However, embryos isolated from XO mothers may be delayed in development (Burgoyne & Biggers, 1976). For the experiment shown in Fig. 3 we were careful to select all embryos at the 8-cell stage, and embryos from XO mothers were collected some hours later than those from XX mothers. Also a recent careful analysis of HPRT and APRT activities throughout development with several time points on the third day of pregnancy has shown that HPRT: APRT ratios do not vary by as much as a factor of two from early to late 8-cell stages (Harper & Monk, unpublished). There is no sign of a developmental lag of JfO-derived embryos at the morula stage (Fig. 4).
In morulae, where the HPRT activity is embryo-coded, the bimodal distribution with respect to HPRT activities strongly suggests that both X chromosomes are active in female embryos. A similar conclusion has been advanced by Adler et al. (1977), and further evidence has been obtained by Kratzer & Gartler (pers. comm.) and Epstein (pers. comm.). In blastocysts the separation of presumptive XX and XY embryos is less than twofold. This result is consistent with the earlier observation (Monk & Kathuria, 1977) that Jf-inactivation has occurred in most, if not all, of the cells of the blastocyst.
ACKNOWLEDGEMENTS
We thank Paul Kratzer, John West and Anne McLaren for advice in the preparation of this manuscript.